Catalytic Vapor-Phase Oxidation of Ethylene. - Industrial

Ind. Eng. Chem. , 1945, 37 (5), pp 432–438. DOI: 10.1021/ie50425a015. Publication Date: May 1945. ACS Legacy Archive. Cite this:Ind. Eng. Chem. 37, ...
3 downloads 0 Views 878KB Size
CATAL TIC VAPOR-PHASE OXIDATION of ETHYLENE E. T. McBee, H. B. Hass, and P. A. Wiseman PURDUE UNIVERSITY AND PURDUE RESEARCH F O U N D A T I O N , LAFAYETTE, IND.

kUOR to 1930 the possibility of A study of the catalytic, vapor-phase air oxidation of ethylene was made using a silver producing ethylene oxide by the catalyst prepared by coating pieces of corundum with silver oxide. The temperature range direct oxidation of ethylene with investigated was between 225' and 395' C. The products isolated were ethylene oxide, air was not indicated in the chemical carbon dioxide, and water. The catalyst was studied to determine the most favorable ilterature. Bone and Wheeler (3) operating conditions for the production of ethylene oxide: A large air-ethylene ratio i s most reported in NO4 that the slow thermal favorable for obtaining both good yields and conversions) the temperature of optimum yields 1)xidation of ethylene Yields no ethyli s somewhat lower'than that of optimum conversion; the temperature of operation of the ene oxide. I n 1922 Blair and Wheeler catalyst must b e raised as the contact time becomes shorter to maintain maximum efficiency) !3) canfirmed this report and stated small additions 06 ethylene dichloride t o the reactants are desirable, large additions are dethat, although acetaldehyde can be leterious; such a catalyst, when deactivated by excess ethylene dichloride, can b e revived by produced in small quantities by such raising the operating temperature and passing the reactants over the catalyst simultaneously. asxidations,no ethyleneoxide is formed. Reyerson and Swearingen (18) stated ID 1928 that the air-oxidation of phase oxidation of ethylene to produce ethylene oxide. During ethylene using metallized silica gels as catalysts gives only carbon the next ten years this chemical process developed rapidly, and dioxide and water. ethylene oxide became one of our most important organic chemiI n 1931, however, Lehner (13) reported that ethylene oxide and cals. Ethylene oxide is now being used in commercial quantities other olefin oxides can be made by the direct noncatalytic oxidafor the production of such vital materials as plasticizers, plastics, tion of olefins using oxygen; Lefort (12) made application in resins, insecticides, coolants, surface-active agents, solvents, and France for a patent on the catalytic oxidation of olefins for the production of olefin oxides. Following these discoveries, a relaexplosives. The development of LE suitable catalyst and the discovery of tivelp large amount of research was done on the catalytic vaporthe best method of applying such a catalyst have been tthe aims of many investigators. Of the numerous materials tested and reported a t least partially successful (4, 6, 9, 10, 22, 16, 181, silver and certain of its compounds seem to be the most desirable catalysts yet found. This metal may be used alone or i n alloys. More commonly, however, a silver compound such as the oxide, nitrate, carb o n a t e , c h l o r i d e , or cyanide is used t o coat a suitable inert carrier. Almost a n y inert material which provides a large surface area can be used as a catalyst carrier. There is evidence, however, that both the chemical nature and the physical state of Figure 1 Apparatus Used in Studying the Reaction for Production of Ethylene O x i d e by Catalytic Air Oxidation of Ethylene the carrier may have an appreciable effect upon 1. A i r cylinder 11. Gases mixed here t 8. Stirrer P. Ethylene cylinder 1% Preheater 19. Direct reading pyrometer the activity of the cata3 4. T-tubes 13. Catalyst chamber PO P4 Stopcocks three-way 5' 6. Glass tubes containins water 14. Thermocouple (bath temperature) 21: P2: Ethylene dxide absorbers lyst (17). Among the 7: Drylng towers 15. Catalyst tube support 23. Pinch clamp carriers which have 16. Thermocouple (catalyst bed tern25. Bubbler 8. A i r flowmeter 26. Gas sampling tube 9. Ethylene flowmeter perature) heen tricd are pumice, I O . Auxiliary inlet 17. Salt bath 37. Mercury leveling bulb

P

~

482

,

INDUSTRIAL A N D ENGINEERING CHEMISTRY

May, 1945

charcoal, activated carbon, silica gel, 7 silica stone, alundum, and corundum r (4, 6, 10, 17, 18). I Of these materials, I it appears that the I 1 most satisfactory I I ones consist of some I form of aluminum I oxide. I I It has been shown I that the addition of I I small quantities of U I modifying agents, -1 I called “promoters”, 3- I often has a desirable I i PREHEATER I effect on the silverLGASES mrm n m 3.CATAL YST MASS type catalysts (4, 4.fNLE T F O 8 14,16). SuchaproTHERIWOCOUPLE I moter is usually an S.THEffMOCOUPLE 6,CROSS SECTlOk oxide or hydroxide CATAL YST TUBE I of an alkaline’earth 7 GASES 1€AYE HERE I I metal.! There seems I L to be little doubt but that catalystsso Figure 2. Reactor Used in Catalytic modified are more Oxidation of Ethvlene rugged and durable, and possess a more active and longer life than similar untreated ones. During the oxidation of ethylene, the products which are obtained depend on the temperature and the manner in which the oxidation is performed. The products formed in the catalytic process are thought to be the result of two competing reactions:

U

-

2CtH4

+ +

0 2 +

1

2CsHdO; AH~BI = -56 kg-cal.

2CzH4 6 0 ~ + 4 c 0 2

+ 4H~0( 9 ) ; AH~DI= -631

433

the barest details concerning the nature of the reaction. Each new investigator in the field, consequently, finds that he must make an independent study of the reaction before any new phase of the problem can be attacked. Undoubtedly, this situation has led to repeated duplioation of effort. The purpose of this study was: first, to establish a definite method for the preparation of a silver catalyst which would be suitable for use in oxidizing ethylene to ethylene oxide with air; second, using the ihformation already available in the patent literature, to learn as much as possible concerning the nature of this reaction; and third, to find a suitable method of testing rapidly any catalyst suspected of possessing desirable properties for this reaction. OPERATION OF APPARATUS

Figure 1 is a drawing of the apparatus used during the experiments. The Pyrex catalyst tube which was found convenient for laboratory use is shown in more detail in Figure 2. The operations performed during a typical experiment were as follows: By regulating carefully the valves attached to the air and ethylene cylinders, 1and 2 (Figure l), the as flow was adjusted so that a small stream of gas issued from t f e ends of T-tubes 3 and 4. Air passed from the former, ethylene from the latter. The approximate desired flow of ases was established by adjusting the depth to which these T-tutes extended into the water contained in tubes 5 and 6.

(1) kg-cal.

(2)

Reaetion 2 is highly exothermic and must be brought under control. Since excessive temperatures favor the formation of carbon dioxide and water rather than ethylene oxide, it is essential that the catalyst be maintained within narrow temperature limits during operation. The more the formation of carbon dioxide and water can be suppressed, the more desirable the process becomes, both from the point of view of economy of operation and of materials. A common method for controlling the temperature of an exothermic reaction is to surround the reactor with a material such as mercury or molten salt. By suitable regulation of this bath, it is possible to maintain the catalyst a t the desired temperature level. Another method widely reported as useful for controlling such a process is the introduction of a diluent along with the reactants. Many such diluents have been tried. Of these, nitrogen, carbon dioxide, air, and steam are the most common (4,6,1%,16,lr). A third and interesting method of bringing the reaction under control is in the use of substances called “anticatalysts” (1, ‘7, 11, 14). An anticatalyst is a specific substance which, when added to the reacting gases in very small amounts, suppresses the formation of carbon dioxide and water and causes the catalyst to be more efficient in the production of the olefin oxide. Ethylene dichloride serves as a good example of such a material. Of the vast amount of work which was performed in the development of the ethylene oxide industry, almost all of it was accomplished in industrial laboratories. No purely academic work seems to have been published on the subject. Although much valuable information is gained by studying the patents, they reveal little concerning the methods of approach to the problem; they tell nothing about the analyses used and give only

temperature of the catalyst mass and the flows of reacting gases were determined ahd recorded. At the end of this interval, stopcock 20 was turned, and the gases were passed through absorbers 22. To determine the amount of ethylene oxide produced, it was necessary t o remove the liquid from absorbers 21 and to titrate the unreacted acid with standard sodium hydroxide solution. The difference between the equivalents of base required for this titration and that required for a blank gave a direct measure of the ethylene oxide produced. Since the time interval during which .the.ethylene oxide was absorbed in 21 for quantitative determination was so short, a second set of absorbers 22 was used to remove all of the ethylene oxide from the exit gases and t o make sure the entire system was sufficiently purged so that a representative gas sample could be collected for analysis. After leaving absorbers 22 the gases normally passed through bubbler 25. By adjusting the stopcock a t the top of gas sampling bulb 26, however, it was posfiible t o collect a gas sample from this exit gas st,ream by the downward displacement of mercury in the bulb. When sufficient time had elapsed for the absorbing system t o be swept out, a representative gas sample was taken in this manner. The gas sample was analyzed on the Orsat apparatus for carbon dioxide, unreacted ethylene, oxygen, and carbon monoxide. No carbon monoxide was detected. Mercury, covered with a thin film of water, was used as the confining liquid in the Orsat so that the gas sample was always saturated with water vapor. Qualitative tests performed on the exit gases indicated that acetaldehyde was not produced in any appreciable quantity. [ PREPARATlONiloFiMATERlALS

SILVERCATALYST. The catalyst was made by coating pieces of corundum of suitable size with silver oxide. This corundum

INDUSTRIAL AND ENGINEERING CHEMISTRY

434

0.10

Vol. 37, No. 5

-

009.

A/ff -ETHYLEN€ RATIO =8.0-/0.3:/ CONTACT T/ME 0.9Q-/.IOSEC. 240

230

240

250

,

,

260

270

I

280

I

250

270 280 JEMPERATURE, "C.

290

260

3w

Figure 5. Yields and Conversions for Experiments Performed at Varied Temperatures but with Approximately Constant AirEthylene Ratio and Contact Time

290

TEMPERATORE, 'C.

Figure 3. Temperature of Optimum Conversion at Constant Contact Time and Air-Ethylene Ratio for a N e w Silver Catalyst

60

40:p*lI-

-

20

TfMPERATURE CONTACT T/M€ '280 I02% -I.Vf SEC

I

t

I

240

250

260

270

280

Tf MPERA TUAE ,

290

300

"C.

Figure 6. Temperatures of Optimum Conversion at Varied Contact Times, Using a Constant Air-Ethylene Ratio

(No, T-61, Tabular alumina) was obtained from the Aluminum Company of America. Seventy grams of corundum of about ,%mesh, 21.98 grams of silver oxide, and 2.24 grams of barium peroxide were placed in a clean beaker. One hundred milliliters of distilled water were added to this mixture. The beaker containing these substances was placed on a steam cone. As the water evaporated, the mixture was stirred vigorously. By the time the water was completely removed, the corundum was well coated with silver oxide. The catalyst \vas then placed in a n oven a t 115' C. and allowed to dry for about 10 hours. This material was ready for immediate use. E ~ H Y L EOXIDE N E ABSORBING SOLUTIOK. The solution which was used for the absorption of ethylene oxide from the gaseous mixtures was prepared by dissolving 920 grams of magnesium chloride (hexahydrate) in 200 ml. of 0.5 N hydrochloric acid. The resulting solution was warmed to 50' C. with vigorous stirring. On cooling, the volume was made up to 1 liter. If solutions

stronger than 0.1 N hydrochloric acid were desired, more acid was added in the proper proportions (16). ETHYLEKE. During these experiments commercial ethylene was used. Samples from the cylinder were analyzed on the Orsat apparatus and were found to contain material, of which 97.57, was absorbed by fuming sulfuric acid. MODIFIED METHYL ORANGE INDICATOR. The modified methyl orange indicator used in these experiments was prepared by the method of Hickman and Linstead (8): 2 grams of methyl orange indicator were dissolved in 500 ml. of water, and this solution was added to an equal volume of 95% ethanol containing 2.8 grams of Xylene Cyano1 FF. CHECKING THE CATALYST

After the catalyst was prepared and placed in the catalyst tube, i t was necessary to test its activity for the production of ethylene

INDUSTRIAL AND ENGINEERING CHEMISTRY

May, 1945

I

-

oxide and to establish a suitable temperature range for using this material in the initial experiments. For this study the elaborate system for handling the exit gases was replaced with a single absorption tube. This tube was partially filled with a measured volume of 0.1 N hydrochloric acid solution saturated with magnesium chloride. A drop of modified methyl orange was added to the solution as an indicator. The ethylene oxide in the exit gases and the hydrochloric acid in the absorbing solution combined to form ethylene chlorohydrin. As a result, the p H of the absorbisg solution increased and the modified methyl orange indicator changed color from orange, to lavender, to colorless, to green. The colorless stage was taken as the end point. A flow of air and ethylene in the ratio of 10.5 to 1 was established over the catalyst bed, which was initially at 225" C. The operator observed the time interval necessary for sufficient ethylene oxide to be produced to cause the indicator to undergo the changes mentioned. Keeping the air-ethylene ratio and the contact time approximately constant, the temperature of the catalyst mass was gradually raised. The time for a similar volume of the absorbing solution to become colorless was recorded at various temperature intervals. The observations from these experiments are given in Table I.

Table

I. Determination of Suitable Operating Temperature for N e w Silver Catalyst

T:mp.,

C. 225 240 245 250 254 260 266 271 280 291

Air Liter/Min. 0.445 0,445 0.445 0.445 0.445 0,445 0.445 0.445 0.445 0.445

Ethylene Liter/Mih. 0,0423 0.0423 0 0423 0,0423 0.0423 0.0423 0.0423 0,0423 0.0423 0.0423

Time T. Min. t o Decolorize 25 Ml. of Absorbing Soln. 14.80 12.50 11.58 11.17 10.67 9.95 9.72 9.88 10.40 13.50

1 T

0.0676 0.0800 0,0865 0.0895 0.0937 0,1005 0.1028 0.1012 0.0962 0.0740

The time necessary for the color change to occur in the indicator is a measure of the ethylene oxide being produced. By plotting the reciprocals of the times as ordinates and the temperatures as abscissas, the curve of Figure 3 was obtained. The temperature at which the maximum occurs in this curve (about 268" C.) was the temperature of optimum conversion for this particular air-ethylene ratio, catalyst, and contact time. OPTIMUM AIR-ETHYLENE RATIO

A series of experiments was performed at approximately constant temperature (268' C.) and contact time (1.04 to 1.13 seconds). The air-ethylene ratios were varied, however, from 3.47:l to 17.5:l. The purpose of this study was to determine the most favorable air-ethylene ratio for such oxidations. The data from these experiments are given in Table 11.

Table

435

'

260

AlR-ETHYLEIIE RAT/O=

coNncT

I

r i w , SECOUDS

Figure 7. Relation between Temperature of Optimum Conversion and Contact Time in Experiments Performed with a Constant Air-Ethylene Ratio

Using the observations from these experiments, the yields and conversions were calculated. Yield represents the percentage of reacting ethylene converted to ethylene oxide. Conversion represents the percentage of total ethylene introduced which was converted to ethylene oxide. I n Figure 4A the values for the yields and conversions are plotted against air-ethylene ratio. The volumes of gases introduced as shown in the tables are the volumes measured a t standard temperature and pressure. The contact times were calculated on the basis that the gases are a t one atmosphere pressure and a t the temperature shown for the reactor. Similar quantitative experiments were performed about 2 weeks later to determine the effect of aging and continued use on the activity of the catalyst. The observations from these experiments are given in Table 11, and Figure 4B is a graphic presentation of the calculated yields and conversions. After approximately a month of almost daily use, the silver catalyst was rechecked to determine if 268' C. still represented a suitable reaction temperature. It was found that the temperature for optimum conversion had changed from around 268" C. to approximately 280' C. It now seemed desirable to know if the yields and conversions, obtained with the new catalyst at 268' C., could be duplicated by the month-old catalyst if it were operated a t 280' C. A series of quantitative experiments was performed. Table I1 and Figure 4C show the results of these experiments. From these various experiments, i t may be concluded that: First, a high air-ethylene ratio is most favorable for obtaining both good yields and conversions, as shown by Figure 4. Second, as the catalyst ages, it becomes less and less efficient for the production of ethylene oxide. Third, on continued use, the temperature of operation of the catalyst must be raised if maximum efficiency for the production of ethylene oxide is to be maintained.

II. Effect of Air-Ethylene Ratio on Production of Ethylene O x i d e

New Catalyst (268' C.) Experiment No. . 1 2 3 4 5 Air" liter/min 0.414 0.362 0.371 0.346 0.314 E t h 'lenea liter/min. 0.0237 0.0405 0.0641 0.07840.0906 Air,&*& &ti0 17.5 8.95 5.80 4.41 3.47 Contact time, sec. 1.04 1.13 1.05 1.07 1.12 Time, nun. 10 5 5 5 6 Ethylene oxide pro5.06 3.95 4.69 3.43 3.06 duced, millimoles Analysis of exit gases, ml. 97.5 100 2 99.2 99.6 99.0 Sample Coz removed 3.6 8.0 9.6 11.8 10.7 CzHa removed 0.1 2.0 6.4 10.2 15.3 Oz removed 13.3 7.3 3.0 0.0 0.1 At standard temperature and pressure.

9'

2-Week-Old Catalyst (268"C.) 1 2 3 0.400 0.365 0.340 0.0340 0.0436 0.0657 11.8 8.37 5.18 1.05 1.11 1.12 10 6 5

Month-Old Catalyst (2800 C.) 1 2 3 4 0.410 0.393 0.371 0.358 0.0220 0.0436 0.0588 0.0794 18.7 9.02 6.32 4.51 1.03 1.02 1.04 1.02 10 5 5 10

6.39

3.84

0.73

3.36

96.6 6.2 0.8 9.6

99.4 9.3 1.9 4.9

98.0 13.0 9.9

9W.O 100.0 4.6 12.4 1.0 2.2 13.2 1.2

0.0

2.79 0.397 0.693 97.2 13.4 6.8

97 4 12.6 12.6

0.0

0.1

INDUSTRIAL AND ENGINEERING CHEMISTRY

436 Table

111.

Determination of Temperature of Optimum Conversion and Yield of Ethylene O x i d e Expt. No. 1 2 3 4 Temp., C. 240 273 288 308 Aira l./min 0.391 0.390 0.390 0,390 EthileneO, i.(min. 0.0435 0.0435 0.0379 0.0379 Air/CzH4 ratio 9.0 9.0 10.3 10.3 Contact time, see. 1.10 1.04 1.03 0.99 Time, min. 5 5 5 5

Ethylene oxide produced, millimoles Analysis of exit gases, ml. Sample COS removed C ~ Hremoved I Oz removed a .

1.63 99.0 4.6 5.6 12.0

2.90 96.8 7.6 3.1 7.0

2.66 99.3 9.1 1.2 5.7

2.25 98.4 11.2 0.9 3.4

At standard temperature and pressure.

TEMPERATURE OF OPTIMUM CONVERSION

OPTIMUM YIELDOF ETHYLENE OXIDE. A series of quantitative experiments was performed with the air-ethylene ratio (9.O:l to 10.3:l) and contact time (0.99 to 1.10 seconds) maintained approximately constant, but with the operating temperature varied. The purpose was to determine the relation between the temperature of optimum conversion and that of optimum yield. The data are given in Table 111. From these observations, the yields and conversions for the experiments were calculated and are shown in Figure 5. These data show that, for a definite airethylene ratio and contact time, the temperature of optimum yield is about 20’ C. lower than the temperature of optimum conversion.

Vol. 37, NO. 5.

decolorize a definite volume of absorbing solution, containing modified methyl orange indicator, was determined a t various temperature intervals. By plotting the reciprocals of these times against temperature, a curve showing the temperature of optimum conversion for this particular contact time was obtained. By plotting these optimum temperatures against contact times. a curve was obtained which shows the temperature of maximum conversion at any contact time. The contact times studied were 0.56 to 0.60, 1.14 to 1.25, 2.34 t o 2.50, and 4.73 to 5.10 seconds. The observations are given in Table IV. A graphical representation of the data is shown in Figures 6 and 7. The relation between the temperature of optimum yield and the temperature of optimum conversion is shown in Figure 5. This study indicates that, for a definite air-ethylene ratio, as the contact time becomes shorter, the temperature of operation of the catalyst must he raised if maximum conversions are to be obtained. THE CATALYSl

EFFECT OF ETHYLENE DICHLORIDE. There are varying reports in the patent literature concerning the effect of certain “anticatalysts” such as ethylene dichloride on silver catalysts (7, 1 1 , 14). It seemed worth while, therefore, to determine how t h e silver catalyst which was used in the previous experiments would respond to small quantities of ethylene dichloride. The catalyst tested was a little over a month old and had been used almost daily. It was relatively inactive when compared with its original state. A small ethylene dichloride bubbler was attached to the 0x1dation apparatus. The equipment was arranged so that, when

a30

Table

IV. Determination of Effect of Contact Time on Optimum Conversion Temperature

kl 0.28-

P F t

Temp., O C. 262 276 284 290 300

k

i?

8B h

0.20

ETHYLENE DICHLORfDE ADIED, GRAMS

Figure 8. 5 Effect of Small Additions of Ethylene Dichloride to Reactants, Air, and Ethylene Valuer for ordinates are proportional to ethylene oxide produced per unit of time.

236 254 260 269 276

Table

V.

F EFFECT OF CONTACT TIME. Ordinarily a great many quanti-

Temp.,

tative experiments must be performed to establish the relation between contact time and the temperatures of optimum yields and conversions. I n this particular process, however, a rapid and convenient method was available for determining the effect of contact time on the temperature of optimum conversion. The experiments were performed with a constant air-ethylene ratio of 10 to 1. For each contact time, the time necessary to

286 285 284 282 282 282 282 282

c.

Contact Time, Sec. 0.60 0.59 0.58 0.57 0.56 1.25 1.23 1.22 1.20 1.19 1.19 1.17 1.16 1.15 1.14 1.14 2.50 2.47 2.45 2.43 2.41 2.39 2.37 2.34 5.10 4.92 4.87 4.79 4.73

Air Lite;/ Min. 0,693 0.693 0.693 0.693 0.693 0.346 0,346 0.346 0.346 0.346 0.346 0.346 0.346 0.346 0.346 0.346 0.174 0.174 0.174 0.174 0.174 0.174 0.174 0,174 0,086 0.086 0.086 0.086

0.086

Ethylene, Liter/ Min. 0,0693 0.0693 0.0693 0.0693 0.0693 0.0346 0.0346 0.0346 0.0346 0.0346 0.0346 0.0346 0.0346 0.0346 0,0346 0.0346

0.0086 0.0086 0.0086 0.0086 0.0086

Time T,Min. to Decolorize 25 Ml. of Absorbing S o h . 9.01 7.83 7.22 7.23 8.46 7.03 5.93 5.25 4.87 4.52 4.50 4.22 4.22 4.28 4.62 4.85 11.58 9.52 8.98 8.25 7.92 7.82 8.28 8.72 19.08 15.17 15.03 16.58 19.90

2

T

0.1110 0.1276 0.1386 0.1384 0.1181

0.0864 0.1052 0.1113 0,1212 0.1264 0,1280 0,1208 0.1148 0.0525 0.0660 0.0666 0,0604 0.0502

Determination of Effect of Ethylene Dichloride on Catalyst C~HICIZ, Air Gram Liter/Min. 0.436 0.436 0.0078 0.436 0.0117 0.436 0.0157 0.436 0,0196 0.436 0.0235 0.436 0.0274 0.436 0.0353‘

....

Time T,M,in. Ethylene, t o Decolorize Liter/ 25 Ml. of AbMin. sorbing S o h . 12.75 0,0417 0.0417 3.45 3.78 0.0417 4.32 0.0417 4.83 0.0417 6.50 0.0417 8.78 0.0417 11.58 0.0417

0.0785 0.2900 0.2650 0.2320 0.2070 0.1540 0.1140 0.0866

INDUSTRIAL AND ENGINEERING CHEMISTRY

May, 11945

u

580

290

300

TEMPERATURE I 'c.

Figure 9. Effect of Simultaneously Raising the Temperature and Passing Air and Ethylene over a Catalyst Deactivated with Ethylene Dichloride Values for ordinates are pro orllonal

i o elhvlens oxide producecfper unit

of time.

desired, nitrogen could be forced through this bubbler into the gaseous mixture of air and ethylene. A flow ratio of air and ethylene of 10.5 to 1 was established. Then ethylene dichloride was introduced intermittently into the gas stream. The changes which occurred in the catalyst were followed by determining, after each addition, the time required to decolorize a definite volume of 0.1 N hydrochloric acid saturated with magnesium chloride to which a drop of modified methyl orange had been added. The data given in Table V and Figure 8 show that small additions of ethylene dichloride are desirable and increase the quantity of ethylene oxide produced by as much as fourfold. Too much ethylene dichloride, however, is deleterious. Ethylene dichloride was introduced until the catalyst became more inactive-than the original untreated material. The purpose of this deactivation was to determine if a

280

-

I

sj

437

I

AIR-ETHYLENE RATIO; I O : /

$280

?

a2 270 B

t

260

I

2

9

CONTACT TIME,

J

4

Sf CONOS

Figure 11, Relation between Temperature of Optimum Conversion and Contact Time in Ex eriments on a Reactivated Catalyst Using a Constant Air-Ethylene Ratio

P

IR-ETHYLENE RATIO = I O : I

250

I

~

I

2

3

4

5

CONFACT r I M E , SECOND$

Figure 19. Relation between Temperatures of Optimum Conversion and Contact Timer for a Catalyst before Deactivation and after Subsequent Revival

catalyst, so tr&ted, was permanently injured. This was not the case. It was found that the catalyst could be reactivated easily by raising the temperature and passing the reactants over the catalyst simultaneously (Table VI and Figure 9).

Table

VI.

Temp.,

C. 281 282 284 286 288 293 296 300 303

O

Reactivation of Silver Catalyst Deactivated with Excess Ethylene Dichloride Air Litar/Min. 0.436 0.436 0.436 0.436 0.436 0.436 0.436 0.436 0.436

Ethylene Liter/MiA. 0.0417 0.0417 0.0417 0.0417 0.0417 0.0417 0.0417 0.0417 0.0417

Time T,Min. t o Decolorize 26 M1. of Absorbing S o h . 12.92 9.47 5.90 4.60 3.77 3.62 3.60 3.60 3.78

1 T

0.0773 0.1056 0.1896 0.2176 0.2651 0.2840 0.2858 0.2778 0.2646

EFFECTOF CONTACT TIMEON OPTIMUMCONVERSION TEMREACTIVATED CATALYST.From the point of view of application, it was desirable to kn ow how a catalyst, which was deactivated with ethylene dichloride and then revived, should be operated to obtain the best results. Therefore, a temperature-contact time study was made on the reactivated catalyst. The experiments were performed in the same manner as the study of the relation between contact time and temperature of optimum conversion for the untreated catalyst. The contact times were varied from 1.14 to 4.94 seconds and were studied at temperatures ranging from 252" to 293" C. A constant a i r - e t h y l e n e r a t i o of 10 t o 1 w a s maintained. PERATURE WITH

Figure 10. Temperatures of Optimum Conversion at Varied Contact Times for a Reactivated Silver Catalyst, Using a Constant AirEthylene Ratio

438

INDUSTRIAL AND ENGINEERING CHEMISTRY

The data are given in Table VI1 and are shown in graphic form in Figures 10 and 11. It appears that, when using a catalyst which has been deactivated with ethylene dichloride and then revived, the temperature of optimum conversion is always several degrees higher than for the untreated catalyst. The curves of Figures 7 and 11, when shown together as in Figure 12 illustrate this point more clearly. ACKNOWLEDGMENT

Thanks are hereby extended to the Monsanto Chemical Company for deferring the expenses of this investigation as Purdue Research Foundation Fellowship 198 and for their cooperation and interest. Bed, E., U. S. Patent 2,270,780 (Jan. 20, 1942). Blair, E. W., and Wheeler, T. S., J . SOC. Chem. I n d . , 41,

Table VII. Determination Qf Effect of Contact Time on Optimum Conversion Temperature with a Reactivated Catalyst I I

I einp.,

c.

258 2i0

275 282 286 293 254 260

262 267 272 277

284 286 2 52

LITERATURE CITED

303T

(1922); 42, 415T (1923).

Bone, W. A,, and Wheeler, R. V., J . Chem. Soc., 85, 1637 (1904). Carter, R. M., U. S.Patents 2,125,333 (-4ug. 2, 1938): 2,177,361 (Oct. 24, 1939). Carter, R. M.. Ibid., 2,294,383 (Sept. 1, 1942). Francon, J., I h i d . , 2,143,371 (Jan. 10, 1939). Halferdahl, A. C., Natl. Research Council of Canada, private communication. Hickman, K. C. D., and Linstead, R.P., J . Ckem. SOC., 121, 2502 (1922).

Langwell, H., Maddocks. C. B., and Short, J. F., U. S.Patent 2,138,583 (Nov. 29, 1938) 2,142,948 (Jan. 3 , 1939); 2,187,882 (Jan. 23, 1940). Law, G. H., and Chitwood, H. C., Ibid., 2,279,469-70 (April 14, 1942). Lefort, T . E., Ihid., 1,998,878 ( A p d 23, 1935).

Law, G. H., I b i d . ,

Vol. 37, No. 5

258 262 266 271

(13) (14)

Contact Time, soc.

1.22 1.19 1.18 1.16 1.16 1.14

2.44 2.42

2.41

2.38

Ais, Liter,/ Min. 0,346

0.346 0.346

0,346 0.346 0.346 0.174 0.174 0.174

0,174

2.36

0,174

2.30

0.174

2.34 2.32

4.94 4.88 4.84 4.80 4.76

0.174 0.174

0.086 0.086 0.086 0.086

0.086

Ethylene, Liter/ Xin. 0 0 0 0 0 0 0

Time T , Alin. t o Decolorize 25 111. of Ahqorhing S o h .

0346 0346

0346 0346

0346 0346 0174

0.0174 n.oii4 0.0174

0,0174 0.0174

n 0174 0.0174

0.0086 0.0086

0,0086

0.0086 0.0086

2.

5.02

2' 0.1560 n. 1990

4 77

o.zino

6.40

4.80 4.72

4.97 10.35 9.30 8.75 8.38 8.22 8.58 9.13 9.53 14.92 14.50 14.37 14.72 16.88

0.2082 o 2120 0 2012

0,0967 0,1070

0.1142 0.1193 0.1218 0.1166 0.1096 n. 1050 0.0672 0.0689 0.0697 0.0679 0,0630

Lehner, S., Ibid., 1,995,991 (March 26, 1935). McNamee, R. W., and Blair, C. >I., Ibid., 2,238,474 (April

15,

1941).

(15) Monsanto Chemical Co., private communication. Peski, ,4.J., van, U. S.Patent 2,040,782 (May 12, 1936). Reyerson, L. H., and Oppenheimer, H.. J . Phus. Chem., 48, 2901

(16) (17)

(1944). (18)

Reyerson, L. H., and Swearingen, L. E., J . Am. Chem.

Soc., 5(9,

2872 (1928).

THISpaper contains material abstracted from a thesis submitted by P. A. Wiseman t o the faculty of Purdue University in partial fulfillment of the requirements for the degree of dootor of philosophy.

Oxidation of Chlorinated Hydrocarbons to

ALEIC

and

FUMARIC ACIDS W. L. Faith

K A N S A S STATE C O L L E G E , M A N H A T T A N , K A N S . Present address, State University of Iowa, low0 City, Iowa.

I

PIT CERTAIK instances polychlorinated aliphatic hydrocar-

bons are industrial by-products of comparatively low value. Accordingly such polychlorides may be available as cheap raw materials for chemical processes. This paper describes a process by which chlorinated hydrocarbons may be converted to more valuable products (Le., maleic and fumaric acids) by simultaneous catalytic vapor-phase oxidation and dehydrochlorination. Hydrochloric acid is a by-product of the reaction. Because of its widespread use, the catalyst chosen for this reaction was vanadium pentoxide deposited on various carriers. I n order to determine the breadth of the reaction, numerous runs were made on various halogenated hydrocarbons in glass laboratory apparatus previously described by the author (4). This was followed by a, series of runs on dichloropentane in larger-scale metal apparatus to obtain data on the effect of temperature and space velocity. This apparatus has also been described previously (5'). The vanadium pentoxide catalyst was prepared as follows: 80 ml. of aluminum pellets (made by compressing aluminum X 1/8 inch) or Alfrax pellets turnings to pellets approximately (3/16 X s/Io inch) were washed with strong oxalic acid solution, rinsed, and just covered with distilled water in an evaporating

dish. Five grams of C.P. ammonium metavanadate were dis. acid were solved in t'he heated water, and 5 grams of c . ~ oxalic then added. This mixture was evaporated to dryness with constant stirring. The resulting catalyst consisted of a firmly adhering precipitate on the carrier. The catalyst was then placed in t,he converter and air was passed over it a t a temperature between 350" and 400" C. for about 6 hours. The resulting catalyst contained about 12.4% vanadium oxides by weight; the carriers were practically free of alkali (Alfrax pellets contained less than 0.0002% iXa20). PRELIMINARY RUNS

Seven different halogenated hydrocarbons were used as raw materials in order to determine the breadt,h of the reaction. These raw materials mere n-butyl chloride, n-butyl bromide, 1,Zdichlorobutane, mixed amyl chlorides, dichloropentane, polychloropentanes, n-hexyl chloride, and "keryl chloride". With the exception of n-butyl bromide, which n-as an Eastman practical grade, all materials were products of Sharples Chemicals, Inc. Of these, the 1,2-dichlorobutane and n-hexyl chloride were comparatively pure. The others were commercial grades with the following specifications :